neon-postgres
Pass
Audited by Gen Agent Trust Hub on Nov 6, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONCREDENTIALS_UNSAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill recommends installing several Node.js packages including @neon/config, @neon/ai-sdk-provider, pg, ai, dotenv, and openai. These are either vendor-owned packages or well-known, established libraries for database connectivity and AI development.
- [COMMAND_EXECUTION]: The skill instructs the agent to use the neon CLI for database operations, including project management, branch creation, and compute updates. This is standard functionality for the vendor's official management tool.
- [CREDENTIALS_UNSAFE]: The skill manages database credentials by reading from and writing to the .env file and checking environment variables for DATABASE_URL. This is a standard and recommended practice for secret management in development environments.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied SQL, schema definitions, and environment configurations. Ingestion points: Environment variables, .env files, and user-provided code or SQL strings. Boundary markers: None explicitly defined in the instructions for prompt interpolation. Capability inventory: Execution of shell commands via the neon CLI, file writing to .env, and SQL execution through database drivers. Sanitization: The skill explicitly recommends best practices such as using parameterized queries and ORM-based migration management to prevent injection.
Audit Metadata